Abstract
We set up a model in which the residents of two neighboring municipalities use the services provided by public infrastructures located in both jurisdictions. The outcome is that municipalities strategically interact when investing in infrastructures, with the small municipality reacting more to the expenditure of its neighbor than the big one. This theoretical prediction is tested by estimating the determinants of the stock of public infrastructures of the municipalities belonging to the Autonomous Province of Trento in Italy. By introducing the classical spatial lag-error component, we find that municipalities positively react to an increase in infrastructures by their neighbors, but the effect vanishes above a given population threshold. Such a result is confirmed when we exploit the exogenous variation in the neighbors’ stock of infrastructures induced by a strong flood that occurred in the Province of Trento in 2000.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
On the other hand, one of the main costs of large jurisdictions is that associated with the need to tailor public services to the increasingly heterogeneous preferences of the citizens, and of gathering information about these preferences. Recently, the empirical literature has tried to disentangle benefits and costs of large jurisdictions by examining the impact of municipal mergers on local fiscal policies. The evidence is mixed. Some studies (see, among others, Reingewertz 2012; Blesse and Baskaran 2016) find that municipal expenditure declines after merger, hinting at the presence of economies of scale. Other studies find no effect of mergers on municipal expenditure (see, among others, Moisio and Uusitalo 2013; Allers and Geertsema 2014). Finally, some studies find that voluntary municipal mergers increase local expenditure, because the merging communities have incentives to free ride on the common pool created by the merger (Saarimaa and Tukiainen 2015).
The fact of limiting the analysis to the case of only two jurisdictions obviously implies that each one of them is ‘the neighbor’ of the other one. We adopt such a simplified setup for analytical convenience. A richer, but also more complex, specification is that of the ‘circular region’, a formalization akin to that used in spatial models of product differentiation, in which the local jurisdictions are located along a circle, so that each one of them has two neighbors, one at its left and one at its right of the regional territory (see Solé-Ollé 2006, for an application of such a type of framework).
Heterogeneity between jurisdictions in terms of the preference parameters \(\alpha _{i}\) and \(\beta _{i}\) can be due to geographical factors, demographic factors (e.g., the share of elderly in total population), characteristics of the local economy, and so on.
In line with the prevalent literature, we assume that the spillover is automatically determined by the provision of local infrastructures. It is possible to extend our framework to the more realistic case in which the effective level of enjoyment depends on usage levels, endogenously chosen by individuals of the two jurisdictions.
Most models analyzing local public goods spillovers assume that the total amount of public goods enjoyed by the residents of any given jurisdiction is equal to a weighted sum of the ‘home’ and the ‘neighbor’ public goods supplies, which implies that the public goods provided by different jurisdictions are perfect substitutes (in our model, this case is obtained by setting \(\phi =-1\)). The more general functional form of the utility function given in Eq. (1) is widely used in oligopolistic models with product differentiation (see, e.g., Singh and Vives 1984).
Note that the specification given in Eq. (2), by which the stock variable \(z_{1}\) is a function of the flow variable \(E_{i}\), implicitly assumes full depreciation of the expenditure in infrastructures within the time period. That is, we consider a simple static model instead of a more complex dynamic framework.
Although no analytical solutions are available for the general case in which \(r\in (0,1)\) and \(\theta <1\), by means of numerical simulations, assuming again \(\psi (N_{j},\theta )=\theta N_{j}\), it is possible to generalize the results of Proposition 1. In particular, for r sufficiently close to zero, the slope of the best response function is, in absolute value, decreasing in \(N_{1}\), for given \(N_{2}\). Instead, for r sufficiently close to 1, it is decreasing in \(N_{1}\) provided that \(N_{1}\) is above a given threshold \(\bar{N}\), with \(\bar{N}<N_{2}\). Details of the simulations are available upon request. We also note that in our model the link between size of the jurisdiction’s population and size of its reaction to the expenditure by its neighbor does not depend on agglomeration economies of the type studied by, e.g., Jofre-Monseny (2013) and Luthi and Schmidheiny (2014).
In Italy there are five Autonomous Regions (Sicily and Sardinia, which are insular territories, and Valle d’Aosta, Trentino Alto-Adige and Friuli Venezia-Giulia, which are northern boundary territories), and two Autonomous Provinces (Trento, with Italian as official language, and Bolzano, with German as official language, making up the Trentino Alto-Adige Region).
For the period covered by our study, the main local tax at the municipal level is Imposta Comunale sugli Immobili (ICI), which is based on the cadastral value of real estate and on the market value of building lots. Minor taxes include a surcharge on the personal income tax and a surcharge on the tax on electricity consumption. User charges include waste collection and fees for public services such as public transport, nursery schools and so on.
We use the 2007 base year deflator for gross fixed capital formation computed by the “Autoritá per l’Energia” (www.autorita.energia.it).
Starting from Gramlich (1977), there is a well-known literature on the effects of grants on public expenditure, usually finding that grants can stimulate government expenditures more than monetary transfers to individuals of the same amount—the so-called flypaper effect, whereby a quota of the federal money sticks to the public sector instead of being distributed to citizens. Interestingly, Wyckoff (1991) finds that capital expenditures are particularly sensitive to grants.
Data on general transfers cover the period 1991–2007 and account for about 60% of capital expenditures of all municipalities. Specific transfers cover only the period 2001–2007 since for the period 1991–2000 there is no distinction between the two categories of transfers in budgetary data. For the period 2001–2007, total transfers account for about 68% of capital expenditure, with specific transfers accounting for about 60% of the total.
To compute the per capita value of the 2001–2007 average stock, we divide it by the average population over the same period. To test the robustness of the results, we built several different measures of the capital stock and found no significant changes (the details are available upon request). In particular, we computed the initial capital stock in year 1994 (obtaining a 14-year series) and in year 2006 (obtaining a 2-year series), assuming no depreciation. Then we also considered linear depreciation rates of 2, 3, 4 and 5%, which are in line with those used in similar studies estimating the stock of public infrastructures, such as those carried out by the World Bank (Agénor et al. 2005; Arestoff and Hurlin 2006) and the IMF (Kamps 2006). For the Italian case, Marrocu and Paci (2008) use a 4% depreciation rate to build a measure of the capital stock series for the period 1996–2003 at the regional level.
The years 1991 and 2001 are the census years and 2007 is the last year of the dataset.
The results of the non-spatial regression model are reported in Appendix, Table 7.
In the first step of the procedure, the regression includes only the spatial lag coefficient and those of the other independent variables, which are estimated by the usual two-stage least squares estimator. In the second step, the spatial error coefficient is estimated by GMM, using the residuals obtained from the regression run in the previous step. Finally, in the last step, the estimation of the spatial error coefficient is used to perform a spatial Cochrane–Orcutt transformation of the data (a procedure which adjusts all the variables of the model for the spatial correlation in the error term), and the coefficient of the spatial lag, as well as those of the other independent variables, is again estimated by two-stage least squares. The procedure has been implemented through the Stata command spivreg, developed by Drukker et al. (2013). Applications of this routine can be found, among others, in Monkkonen et al. (2012), Zheng et al. (2014) and Glaeser et al. (2015).
It must be noted that for Trento and Rovereto the combined coefficient is negative and statistically significant because their population is by far larger than that of the other municipalities. If Trento and Rovereto are excluded from the regression, it turns out that the combined coefficient measuring the slope of the reaction function is positive and statistically significant for population levels below 1700 inhabitants, while it is not significantly different from zero for all municipalities above 1700 inhabitants (regression results available upon request).
Again, the combined coefficient is negative and statistically significant above a given population threshold, which in the case of Roads and Transport is equal to 6000 inhabitants, and hence it includes the 12 biggest municipalities. The reason being also in this case that the population of these municipalities is, by far, larger than that of the other municipalities.
The main reason is that the characteristics of neighboring municipalities can have a direct effect on the infrastructures spending of a given municipality. In addition, infrastructures spending of neighboring municipalities can influence their own characteristics. Therefore, the characteristics of the neighbors cannot be used as instruments.
The estimates of losses are reported in Law 5/2001, n.1 of the Province of Trento (http://www.consiglio.provincia.tn.it/leggi-e-archivi/codice-provinciale/archivio/Pages/Legge%20provinciale%205%20febbraio%202001,%20n.%201_6926.aspx).
For a detailed account of the physical damages, see the report by Protezione Civile at http://www.protezionecivile.tn.it/binary/pat_protezione_civile/alluvionipiene/UNIONEDOC.1304668153.
Data on the millimeters of rainfall can be found at http://www.meteotrentino.it/dati-meteo/stazioni/elenco-staz-hydstra.aspx?ID=151.
The variables included, expressed in difference from 2000 to 2001, are: grants, population, surface (surface area divided by population; i.e., the inverse of population density), aged, children, houses, total employees and local units.
A possible explanation is that the higher was the intensity of the flood, the higher were the damages. Clearly, while damages in infrastructures of a small entity can be recovered in a short time, the process of recovery spending for severe damages usually takes several years. Therefore, it is plausible that the increase in spending on infrastructures immediately after the flood was lower in a municipalities heavily hit by the event with respect to other municipalities. The difference, from 2001 to 2002 (2 years after the flood), in the per-capita spending on infrastructures for the municipalities hit by the flood is, on average, equal to 27.78 euros in per-capita terms, while the corresponding figure for the municipalities not hit by the flood is, on average, equal to −12.24, suggesting that 2 years after the flood the former municipalities spent more on infrastructures than the latter. Moreover, we checked whether this increase in per-capita spending on infrastructures between 2001 and 2002 was driven by those municipalities more severely affected by the flood. In particular, we used the intensity of the flood to classify municipalities into two groups: a) those strongly affected by the flood and b) those slightly affected by the flood, with the former group containing those municipalities for which the intensity of the flood is above the value identifying the third quartile (483 millimeters). Interestingly, we found that the increase in the per-capita spending on infrastructures between 2001 and 2002 is more marked for those municipalities strongly affected by the flood (76.68 per capita euros) than for those slightly affected (8.22 per capita euros).
References
Agénor, P. R., Nabli, M. K., & Yousef, T. M. (2005). Public infrastructure and private investment in the Middle East and North Africa. World Bank Policy Research Working Paper, No. 3661.
Allers, M. A., & Geertsema, J. B. (2014). The effects of local government amalgamation on public spending and service levels Evidence from 15 years of municipal boundary reform. Mimeo (University of Groningen).
Anselin, L. (1988). Spatial econometrics: Methods and models. Boston: Kluwer Academic Publisher.
Arestoff, F., & Hurlin, C. (2006). Estimates of government net capital stocks for 26 developing countries, 1970–2002. World Bank Policy Research Working Paper, No. 3858.
Baicker, K. (2005). The spillover effects of state spending. Journal of Public Economics, 89(2), 529–544.
Baskaran, T. (2014). Identifying local tax mimicking with administrative borders and policy reform. Journal of Public Economics, 118, 41–51.
Bergstrom, T. C., & Goodman, R. P. (1973). Private demands for public goods. The American Economic Review, 63(3), 280–296.
Blesse, S., & Baskaran, T. (2016). Do municipal mergers reduce costs? Evidence from a German federal state. Regional Science and Urban Economics, 59, 54–74.
Borcherding, T. E., & Deacon, R. T. (1972). The demand for the services of non-federal governments. The American Economic Review, 62(5), 891–901.
Brueckner, J. K. (2003). Strategic interaction among governments: An overview of empirical studies. International Regional Science Review, 26(2), 175–188.
Case, A. C., & Rosen, H. S. (1993). Budget spillovers and fiscal policy interdependence: Evidence from the States. Journal of Public Economics, 52(3), 285–307.
Cliff, A., & Ord, J. (1973). Spatial Autocorrelation. London: Pion.
Cliff, A., & Ord, J. (1981). Spatial processes. Models and applications. London: Pion.
Cohen, J. P., & Paul, C. J. M. (2004). Public infrastructure investment, interstate spatial spillovers, and manufacturing costs. Review of Economics and Statistics, 86, 551–560.
Costa, H., Veiga, L. G., & Portela, M. (2015). Interactions in local governments’ spending decisions: Evidence from Portugal. Regional Studies, 49, 1441–1456.
Cremer, H., Marchand, M., & Pestieau, P. (1997). Investment in local public services: Nash equilibrium and social optimum. Journal of Public Economics, 65, 23–35.
Dahlby, B. (1996). Fiscal externalities and the design of intergovernmental grants. International Tax and Public Finance, 3, 397–412.
Drukker, D., Egger, P., & Prucha, I. R. (2010). On two-step estimation of a spatial autoregressive model with autoregressive disturbances and endogenous regressors. Technical Report, Department of Economics, University of Maryland.
Drukker, D., Prucha, I. R., & Raciborski, R. (2013). A command for estimating spatial-autoregressive models with spatial-autoregressive disturbances and additional endogenous variables. The Stata Journal, 13, 287–301.
Gibbons, S., & Overman, H. (2012). Mostly pointless spatial econometrics? Journal of Regional Science, 52, 172–191.
Glaeser, E. L., Kerr, S. P., & Kerr, W. R. (2015). Entrepreneurship and urban growth: An empirical assessment with historical mines. The Review of Economics and Statistics, 97, 498–520.
Goldsmith, R. W. (1951). A perpetual inventory of national wealth. In M. R. Gainsburgh (Ed.), Studies in income and wealth 14 (pp. 5–74). Princeton: Princeton Univeristy Press.
Gramlich, E. M. (1977). Intergovernmental grants: A review of the empirical literature. In W. E. Oates (Ed.), The political economy of fiscal federalism (pp. 219–239). Lexington, MA: Lexington Books.
Haughwout, A. F. (2002). Public infrastructure investments, productivity and welfare in fixed geographic areas. Journal of Public Economics, 83(3), 405–428.
Jofre-Monseny, J. (2013). Is agglomeration taxable? Journal of Economic Geography, 13, 177–201.
Kamps, C. (2006). New estimates of government net capital stocks for 22 OECD countries, 1960–2001. IMF Staff Papers, 53, 120–150.
Kelejian, H., & Prucha, I. R. (1998). A generalized spatial two-stage least squares procedure for estimating a spatial autoregressive model with autoregressive disturbances. The Journal of Real Estate Finance and Economics, 17, 99–121.
LeSage, J., & Pace, R. K. (2009). Introduction to spatial econometrics. Boca Raton: CRC Press, Taylor & Francis Group.
Lyytikäinen, T. (2012). Tax competition among local governments: Evidence from a property tax reform in Finland. Journal of Public Economics, 96, 584–595.
Luthi, E., & Schmidheiny, K. (2014). The effect of agglomeration size on local taxes. Journal of Economic Geography, 14, 265–287.
Marrocu, E., & Paci, R. (2008). The effects of public capital on the productivity of the Italian regions. Applied Economics, 42, 989–1002.
Meinen, G., Verbiest, P., & de Wolf, P. P. (1998). Perpetual inventory method: Service lives, discard patterns and depreciation methods. Technical report, Statistics Netherlands.
Moisio, A., & Uusitalo, R. (2013). The impact of municipality mergers on local public expenditures in Finland. Public Finance and Management, 13, 148–166.
Monkkonen, P., Wong, K., & Begley, J. (2012). Economic restructuring, urban growth, and short-term trading: The spatial dynamics of the Hong Kong housing market, 1992–2008. Regional Science and Urban Economics, 42, 396–406.
Murdoch, J., Rahmatian, M., & Thayer, M. (1993). A spatially autoregressive median voter model of recreation expenditures. Public Finance Quarterly, 21, 334–350.
Parchet, R. (2014). Are local tax rates strategic complements or strategic substitutes? IdEP Economic Papers, 2014/07.
Reingewertz, Y. (2012). Do municipal amalgamation work? Evidence from municipalities in Israel. Journal of Urban Economics, 72, 240–251.
Revelli, F. (2002). Testing the taxmimicking versus expenditure spill-over hypotheses using English data. Applied Economics, 34, 1723–1731.
Saarimaa, T., & Tukiainen, J. (2015). Common pool problems in voluntary municipal mergers. European Journal of Political Economy, 38, 140–152.
Singh, N., & Vives, X. (1984). Price and quantity competition in a differentiated duopoly. The RAND Journal of Economics, 15(4), 546–554.
Solé-Ollé, A. (2006). Expenditure spillovers and fiscal interactions: Empirical evidence from local governments in Spain. Journal of Urban Economics, 59, 32–53.
Wyckoff, P. G. (1991). The elusive flypaper effect. Journal of Urban Economics, 30, 310–328.
Zheng, S., Sun, W., & Wang, R. (2014). Land supply and capitalization of pubic goods in housing prices: Evidence from Beijing. Journal of Regional Science, 54, 550–568.
Acknowledgements
We wish to thank the Autonomous Province of Trento for kindly providing the data, and Piero Giarda, Ian Preston two anonymous referees and seminar participants in Siegen, Barcelona and Lecce, at the 2014 Gerard-Varet Conference, Aix-en-Provence, the 2014 APET Conference, Seattle, the 2014 IIPF Conference, Lugano, for useful comments. We also thank Riccardo Secomandi for excellent assistance with the data. Massimiliano Ferraresi gratefully acknowledges funding from the University of Ferrara; Umberto Galmarini and Leonzio Rizzo thankfully acknowledge financial support from the Spanish Ministry of Economy and Competitiveness (ECO2012-37873) and Leonzio Rizzo acknowledges financial support also from “Bando FIR”.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Ferraresi, M., Galmarini, U. & Rizzo, L. Infrastructure spillovers and strategic interaction: does the size matter?. Int Tax Public Finance 25, 240–272 (2018). https://doi.org/10.1007/s10797-017-9449-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10797-017-9449-0
Keywords
- Local public goods spillovers
- Spatial interactions
- Size of local jurisdictions
- Natural disaster
- Internal instruments
- External instruments